Lecture materials (JuPyter notebooks):
github-link for the conference with all the talks and the jupyter notebooks:
https://github.com/Depart-de-Sentier/brightcon-2023-talks
Collection
• Hestia has above 60 models / transformations. 650 Studies input (1 more every week), and possibly 40k aggregations (5-15 countries)
Brightway also started working on linking with Hestia. The full Git repository is available at https://github.com/brightway-lca/bw_hestia_bridge We may simply develop this code rather than start our own (Hestia does not have process and flow relationships) → https://docs.brightway.dev/projects/hestiabridge/format.html
Synchronize brightway (non-disctructively) add a tricker to the sql database that track the changes to be a table. From there we should be able to import just the relevant parts (chris suggested it to commit it to bw-data)
• Mike’s idea: Import brightway into EDB more gracefully. have inotify check whether something has changed in the brightway database. (Holger: There are multiple files in the Brigthway database. Does this cause any complications?)
https://app.asana.com/0/1202216409500980/1205546409181590/f
• Hestia Emissions Models: they are modular
https://gitlab.com/hestia-earth/hestia-engine-models/-/issues/?label_name[]=Brightcon 2023
→ very specific documentation of the node looks like before and after. We should add a readme, add a subfolder (REQUIREMENTS
/ RETURNS
). We can modify the syntax for the requirements here (that takes a graph into account!)
→ Did it complete, has it failed (the sequence): Hestia’s models are completely modular and the code does not care about the sequence. Therefore, there is no information in the documentation about the sequence. We should have in our documentation the dependencies between different gap-filling models.
→ we add/remove nodes. Screen the whole graph. They get a big thing json.
→ Diverge:
→ if requirements are not fulfilled, then it skips the models (really hard to trace).
→ Chris said: Banana, with uuid goes into the model. And comes out with the same uuid and name. Makes it difficult to put it in a matrix.